Translating Europe Forum 2015 (European Commission, Brussels, 29-30 October 2015)
Session “The role of translation for multilingualism” (30 oct., 10.00; ~100 attendants)
Timed participation list of the debate on language technologies during the session
Moderator: Konrad Fuhrmann (EC, DG Education and Culture)
- Panagiotis Alevantis (EC, DG Translation)
- Zoe Moores (UK, translator, expert in respeaking)
- Judit Sereg (Hungary, translator, expert in audiovisual translation)
Participating from the audience: Gonçal Garcés Díaz-Munío and Jorge Civera Saiz (MLLP research group, Universitat Politècnica de València)
Watch the session’s video at https://scic.ec.europa.eu/
streaming/translating-europe- forum-2015-jenk-30-10 (debate on language technologies from 39.40; available in English, French and German)
Timed participation list:
- 18.00 – 22.30 Panagiotis Alevantis: Comments on multilingualism and translation in EU policies towards 2020.
- 19.47 – 22.30 Panagiotis Alevantis: Mention of the European Commission’s commitment on language technologies
- 39.40 Gonçal Garcés (MLLP): Participation in the session’s open discussion about how this is an excellent time to apply language technologies in more and different ways.
- 42.12 Konrad Fuhrmann and Gonçal Garcés (MLLP): The moderator enquires whether the final quality of the texts is guaranteed through human intervention. Answer (in short): yes.
- 43.12 Konrad Fuhrmann: The moderator introduces the topic of the situation of language technologies as regards less-spoken languages.
- 43:41 Judit Sereg: Expresses opinion that there is still no machine translation that works well with Hungarian.
- 44.50 Panagiotis Alevantis: On how large enterprises (Microsoft, Google) accumulate data and technology in a way that makes competition difficult in the field of language technologies.
- 48.20 Zoe Moores: On the lack of support for Welsh in commercial speech recognition software.
- 48.56 Gonçal Garcés (MLLP): On the importance of having adequate data (corpora); on how researchers don’t always have access to all corpora; and how, nevertheless, for less-spoken languages researchers are managing to obtain better and better results even with smaller corpora.